See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315175882

Crossover Operators in Genetic Algorithms: A Review

Article in International Journal of Computer Applications · March 2017

DOI: 10.5120/ijca2017913370

CITATIONS READS

99 11,614

2 authors:

Padmavathi Kora Priyanka Yadlapalli

105 PUBLICATIONS 1,472 CITATIONS

Gokaraju Rangaraju Institute of Engineering & Technology
19 PUBLICATIONS 167 CITATIONS
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

journal View project

All content following this page was uploaded by Padmavathi Kora on 23 September 2017.

The user has requested enhancement of the downloaded file.

 International Journal of Computer Applications (0975 – 8887)
Volume 162 – No 10, March 2017

Crossover Operators in Genetic Algorithms: A Review

Padmavathi Kora Priyanka Yadlapalli
Gokaraju Rangaraju Gokaraju Rangaraju
Institute of Engineering & Institute of Engineering &
Technology, Hyderabad Technology, Hyderabad

from the second parent. In genetic algorithms, chromosomes

ABSTRACT denoted as linear strings of symbols (C Ryan et.al, 2000).
Genetic Algorithms are the population based search and
optimization technique that mimic the process of natural It is very important that in order to enhance the
evolution. Genetic algorithms are very effective way of functionality of genetic algorithm problem there is a specific
finding a very effective way of quickly finding a reasonable requisition of crossover. After reproduction or after the so
solution to a complex problem. Performance of genetic called selection process, the population comprises of better
algorithms mainly depends on type of genetic operators which individuals. Reproduction or selection produces same
involve crossover and mutation operators. Different crossover individual but the progeny are improved when compared to
and mutation operators exist to solve the problem that the older ones. The main feature of genetic algorithms is to
involves large population size. Example of such a problem is merge both exploration and exploitation in an ideal way. In
travelling sales man problem, which is having a large set of reality, we are accustomed to the fact that it is the population
solution. In this paper we will discuss different crossover size that has an impact on the performance of genetic
operators that help in solving the problem. algorithm which results in the problem of genetic drift which
is rampant mostly in the case of multimodal search space. By
Keywords initiating a local search method within the genetic operators
Genetic Algorithm; Mutation; crossover; Selection; travelling we can produce new genes that can resolve the problem of
salesman problem genetic drift and accelerate the search towards global optima
(W.E. Hart et.al, 1994). When a genetic algorithm with a local
1. INTRODUCTION search method is combined a hybrid genetic
Evolutionary algorithms always follow the Darwin concept of algorithm/mimetic algorithm is evolved. Some crossover
“Survival of the fittest” which is extensively used for operators are utilized for exploitation as well as for
optimization problems for nearly four decades (J. Holand exploration. The main emphasis of this paper is to study
et.al, 1975), (Padmavathi Kora et.al, 2016) (Ambika various types of crossover operators
Annavarapu et.al, 2016). Evolutionary algorithms are classic
examples of heuristic search algorithms yet they do not yield 2. TRAVELLING SALESMAN
exact optimal solutions, but will certainly help to find better PROBLEM
optimal solutions when compared to their parallel counterparts The Travelling Salesman Problem (TSP) is one of the most
within less amount of time. Genetic algorithms and widely used and well studied for combinational optimization
Evolutionary programming combined provide an excellent problems. The statement of Travelling Salesman Problem
system for effectively sampling large search areas with basic (TSP) was made by Karl Menger in 1930 whose statement is
technique is widely applicable. Genetic algorithms are by very uncomplicated, but it is one of the most challenging and
nature have adaptive optimization algorithms which resemble popular problems in the field of Operational Research. Large
the process of natural selection and genetics (D.E. Goldberg number of articles and research papers have been written and
et.al, 1989).Both exploitation and exploration techniques are published on this problem. The most common practical
by and large accountable for the performance of genetic interpretation of TSP is that a salesman seeking the shortest
algorithms. Exploitation in fact is defined as usage of already tour through n clients or cities. The Travelling Salesman
existing information to find out the better solution and Problem is one of the famous NP-hard problems, which means
Exploration means to find out a novel and unknown solution that there is no perfect algorithm to solve it in polynomial
in investigation space. The crux of the genetic algorithms time. The minimal time required to obtain optimal solution is
emerges from the capability to combine both exploration and exponential (Ivan Brezina Jr et.al, 2011). The main motive is
exploitation in an ideal way . Biological genetics model is to find the shortest path or the path with least cost. It is very
regarded as one of the best role model of genetic algorithm. complicated as starting and finishing points are same and we
A generic genetic algorithm comprises of the following have to return to the original point after visiting each of the
operations namely: Initialization, Selection, Reproduction and vertex only once. The most prevalent practical applications of
Replacement. Initialization represents creation of initial TSP is that it not only involves regular distribution of goods
population by utilizing suitable encoding scheme. Selection or resources, but finding out the shortest path for costumer
operator chooses the individuals arbitrarily or chosen servicing route, planning bus routes , vehicle routing,
according to their fitness. Crossover and mutation genetic computer wiring, machine sequencing and scheduling,
operators are exclusively involved to strike balance between frequency assignment in communication networks , but also in
exploitation and exploration. During replacement, the old ones the areas that have nothing to do with travel routes.
are succeeded by new offspring’s (Manju Sharma et.al, 2013).
By definition, crossover selects genes from parent
chromosomes and creates a new offspring. The simplest way
is to choose randomly some crossover point, point copy from
a first parent and then everything after a crossover point copy

34
 International Journal of Computer Applications (0975 – 8887)
Volume 162 – No 10, March 2017

3. VARIOUS METHODS USED FOR 2) N Point Crossover

The N-point crossover was first implemented by De Jong in
RESOLVING TSP 1975. It has many cross over sites but rule used is same as we
In 1997, Rong Yang initiated several knowledge- used in single point crossover. In 2-point crossover
augmented genetic operators that guide the genetic algorithm significance of crossover sites is 2. Adding more and more
more directly towards better quality of the population but not crossover sites effects the disruptions of building blocks that
trapped in local optima prematurely. The algorithm uses an sometimes reduce the performance of genetic algorithm. But it
avid crossover operator along with two advanced mutation allows the head and tail section of a chromosome to be
operations based on the 2-opt and 3-opt heuristics (Rong Yang accepted together in the offspring
et.al, 1997). In 2001, Chiung Moon introduced the concept of
topological sort (TS), which is as an or interpreted as ordering 3) Uniform Crossover
of vertices in a directed graph. Moreover, a new crossover Uniform crossover do not fragments the chromosomes for
operation is introduced for the proposed GA (Chiung Moon recombination. Each gene in offspring is created by
et.al, 2001). In 2004, new knowledge based multiple inversion copying it from the parent chosen according to the
operators and neighborhood swapping operator is proposed by corresponding bit in the binary crossover mask of same length
Shubhra Sankar Ray (ShubraShankar et.al, 2001). In 2005, as the length of the parent chromosomes . If the bit in
Lawrence V. Snyder gave a heuristic to resolve the crossover mask is 1, then the resultant gene is copied from the
generalized travelling salesman problem. The approach first parent and if the bit in crossover mask is 0, then the
incorporates a local tour development heuristic into a random- resultant gene is copied from the second parent. A new
key genetic algorithm. The algorithm was very much crossover mask is generated arbitrarily for each pair of parent
successful when tested on a set of 41 standard problems with chromosomes. The quantity of crossover point is not fixed
familiar optimal objective values. In 2005, Milena initially. So, the offspring have a mixture of genes from both
Karova found that genetic algorithm when implemented the parents
gives a maximal approximation of the problem, thus
4) Three Parent Crossover
reforming the acquired solution with genetic operators. In the
In this crossover, three parents are chosen arbitrarily. Each
year 2006, Plamenka Borovska analyzed the efficiency of
gene of the 1st parent is compared with the equivalent gene of
parallel computation of travelling salesman problem applying
the 2nd parent. If both genes are similar, the gene is occupied
the genetic approach on a casual multicomputer cluster .In
for offspring or else the equivalent gene from the 3rd parent is
2007, a two-level genetic algorithm (TLGA) was developed
taken for the offspring. It is mostly used in case of binary
for a case, with special treatment consisting of neither intra-
encoded chromosomes.
cluster paths nor inter- cluster paths, thus realizing an
integrated evolutionary optimization for both levels of the 5) Arithmetic Crossover
CTSP. In 2008, a software system is developed to Arithmetic crossover is used in case of real-value encoding.
determine the ideal route for Travelling Salesman Arithmetic crossover operator linearly combines the two
Problem by employing Genetic Algorithm technique. In 2009, parent chromosomes.Two chromosomes are particular
S.N. Sivanandam presented two approaches i.e Genetic randomly for crossover and create two offspring’s which are
Algorithms and Particle swarm optimization in order to find a linear mixture of their parents
solution to the given objective function by applying different
procedures and computational techniques; as a result of which 6) Partially mapped Crossover
their performance can be evaluated and compared. In 2016, Partially Matched or Mapped Crossover (PMX) is the most
Hybrid Particle Swarm Optimization (PSO) and Bacterial frequently used crossover operator. It was proposed by
foraging algorithm has been initiated for the classification of Goldberg and Lingle for Travelling Salesman Problem In
ECG. Partially Matched Crossover, two chromosomes are
associated and two crossover sites are chosen arbitrarily. The
4. GA METHODOLOGY fraction of chromosomes between the two crossover points
A genetic algorithm defines how an algorithm can generate a gives a corresponding selection that undergoes the crossover
modified copy of itself in order to achieve an objective. The process through position-by- position exchange operations
progeny would match against the criteria to ensure that the (Padmavathi Kora et.al, 2015). PMX tends to respect the
objective would be eventually attained. Genetic algorithm absolute positions
which is inspired by Darwin's theory is all about evolution. It 7) Crossover ORDER (OX)
is not that difficult to program or realize, since they are It was proposed by Davis and also used for chromosomes with
biological based. permutation encoding. The process starts in a way similar to
that of PMX by choosing two crossover points. But in its
5. TYPES OF CROSSOVER place of using point-by-point exchanges as in case of PMX,
OPERATORS order crossover applies sliding motion to fill up the left out
1) Single Point Crossover holes by sending the mapped positions. It copies the portion of
Single point crossover is the most approved crossover which permutation elements between the crossover points from the
is widely in use. A crossover site is randomly selected along cut string directly to the offspring, insertion them in the same
the length of the mated strings and bits which are very next to absolute position (Padmavathi Kora et.al, 2015). OX tends to
the cross-sites are interchanged. When suitable site is chosen respect the relative positions.
then better progeny can be obtained by combining the good
qualities of the parents. If appropriate site is chosen, better 8) Cycle Crossover (CX)
offspring can be obtained by combining good quality parents This crossover is used for chromosomes with permutation
else it harshly hampers string quality. In single point crossover encoding. During recombination in cyclic crossover there is a
the head and tail of one chromosome break up and if both limitation that each gene either comes from the one parent or
head and tail have good genetic material then none of the the other. The fundamental model at the back cycle crossover
offspring will get the both good features directly. is that each allele comes from one parent jointly with its

35
 International Journal of Computer Applications (0975 – 8887)
Volume 162 – No 10, March 2017

position. To construct a cycle of alleles from parent1 we have want to prolong the same process with different other
to start with the first allele of parent1. Then look at the allele crossover and mutation operator to give an improved solution
at the equal position in parent2 and go to the position with the to a TSP.
same allele in Parent1.insert this allele to the cycle and do
again above step until you reach your destination at the first 9. REFERENCES
allele of parent1. Put the alleles of the cycle in the first child [1] Ambika Annavarapu, Padmavathi Kora, "ECG-based
on the positions they have in the first parent and the remaining Atrial Fibrillation detection using different orderings of
alleles of first child come from the second parent along with Conjugate Symmetric–Complex Hadamard Transform,"
their position. Produce next cycle from parent2. International Journal of Cardiovascular Academy,
Elsevier, Aug 2016.
6. SOLVING TSP USING GENETIC
[2] C. Ryan, Automatic re-engineering of software using
ALGORITHM genetic programming, Genetic Programming Series.
A genetic algorithm is popular to find a solution of a Kluwer Academic Publishers, ISBN 0-7923-8653-1,
problem in much less time. Even though it might not find the 2000.
most excellent solution of the problem but it can help in
finding a near ideal solution for a 100 city tour in less than a [3] Chiung Moon, Jongsoo Kim, GyunghyunChoi
minute. The following are some fundamental steps of our ,YoonhoSeo,” An efficient genetic algorithm for the
proposed work traveling salesman problem with precedence
constraints”, European Journal of Operational Research
Encoding: Permutation encoding is most popular way to 140 (2002) 606–617, accepted 28 February 2001
solve the TSP. We characterize cities with an integer value,
and after that we initialize the population. [4] D.E. Goldberg, Genetic algorithms in search,
optimisation, and machine learning, Addison Wesley
• Distance matrix: distance matrix is a Longman, Inc., ISBN 0-201-15767-5, 1989
representation of 𝑁 × 𝑁 matrix of point to point
distances. [5] Ivan Brezina Jr.,ZuzanaCickova, “Solving the
Travelling Salesman Problem using the Ant
• Selection based on fitness function: the fitness colony Optimization”, Management Information
function will be total expenditure of the visit Systems, 2011, Vol. (6), No. (4).
represented by each chromosome. The smaller the
sum, the fitter the solution represented by that [6] J. Holland, Adaptation in natural and artificial systems,
chromosome. University of Michigan Press, Ann Arbor, 1975.

• Generating arbitrary numbers equal to population [7] Manju Sharma , Novel Knowledge based
size. Best of two routes will be chosen using Selective Tabu Initialization in Genetic algorithm,
tournament selection to relate Mutation. IJARCSSE, Volume 3, Issue 5, May 2013

• Next generation of population size will be [8] Padmavathi Kora, K. Sri Rama Krishna, “Bundle Block
produced. Process will go through predefined Detection using Genetic Neural Network,” International
iterations. System design and Intelligent Applications (INDIA),
AISC, Springer, Jan 2016.
• After the final iteration the smallest number of
distance size will be displayed as result. [9] Padmavathi Kora, and Sri Rama Krishna, “Hybrid
Bacterial Foraging and Particle Swarm Optimization
7. CONCLUSION for detecting Bundle Branch Block,” SpringerPlus,
In this paper, we have discussed eight types of mutations Springer, vol 4, no 1, 481, Sep 2015.
strategy in the genetic algorithm procedure to create new and
[10] Padmavathi Kora, and Sri Ramakrishna Kalva,
better generations. As we learnt that genetic algorithms appear
”Improved Bat algorithm for the detection of myocardial
to be a good solution for TSP, however it very much depends
infarction,” SpringerPlus, Springer, vol 4, no. 1, pp. 1-18,
on the technique the difficulty is encoded and which
Nov 2015.
crossover and mutation methods are used. So our labors will
decide which mutation operator will always gives end result to [11] Rong Yang, “Solving Large Travelling Salesman
undertake most of the real life applications of travelling Problems with Small Populations”. IEEE 1997.
salesman problem.
[12] ShubhraSankar Ray, Sanghamitra Bandyopadhyay and
8. FUTURE SCOPE Sankar K.Pal,” New Operators of Genetic Algorithms
It appears that major dilemma with the genetic algorithm for for Traveling Salesman Problem”, 2000IEEE.
TSP is that to preserve the formation from the parent’s [13] W. E. Hart, “Adaptive global optimization with
chromosomes and still finish up with a permissible tour in the local search, Doctoral diss.”, San Diego, University of
child chromosomes. Possibly a improved crossover or California, 1994.
mutation routine that remain structure from parent
chromosome would give a far better clarification than we have
previously establish for some TSP. As a potential work we

IJCATM : www.ijcaonline.org
36

View publication stats